I actually agree with what Kroah's opinion.
As long as the user-level api remains the same, most developers would
never even realize that the internel kernel api changes.
The internal kernel api must change for there to be any sort of
progress. I think what happened in 2.6.19 (even though breaking several
former fs drivers - i.e. yaffs) was fantastic, considering it allowed
the kernel up to some huge amount of Terabytes of block IO using the new
block IO scheduling api.
I've had to deal with this a couple of times - specifically when writing
modules for use both in 2.4 and 2.6 kernels - although you would only
realize it if you were using internal api's (or udev / devfs)
~/Chris
Curtis Monroe wrote:
> On March 21, 2007 01:29 pm, Jesse Off wrote:
>>>> In other words...Can embedded linux developers hide their close
>> source
>>> divers
>>>> in a custom close source BIOS, that is called from an open
>> source "wrapper
>>>> module"?
>>> No.
>> I disagree. Inevitably what it would come down to is that somebody
>> would have to be able to convince a judge somewhere that the closed
>> source BIOS is a derived work of the GPL'ed kernel. That would be
>> quite a stretch if the "BIOS" uses no GPL'ed functions or header
>> files and can compile and be useful without Linux or an OS at all.
>>
>> I believe Linux has an NDIS wrapper driver to allow closed source
>> Windows network drivers to be used with Linux on the PC. Does this
>> mean that these Windows drivers must now be made GPL? This is really
>> the same thing.
>>
>
> Since close source drivers violate the Linux GPL, distros like Red Hat force
> users to download and install the drivers separately. So technically the user
> is the the one violating the GPL not Red Hat Inc.
>
> The other solution is the binary BLOB where a closed source device driver is
> converted to an unreadable and uneditable open source blob and linked to the
> kernel. Technically its "open source". The legality of blobs is in question
> too.
>
>
>>>> PCs have close source BIOSes.
>>> Some do, some do not.
>> True, 99.9% do, .1% do not. Seriously, this seems a ridiculous
>> statement-- what brand PC currently ships with an open-source BIOS?
>>
>>>> My problem is the SD flash card interface is patented and its
>> governing body
>>>> doesn't allow release of open source drivers for the SD
>> interface.
>>> That's not true, we have SD drivers in the Linux kernel now.
>> They are probably MMC drivers, not full SD. SD cards are backward
>> compatible with MMC. Either that or they were developed illegally or
>> reverse engineered in a country that allows that. SD association
>> licenses are pretty clear. You might be able to get away with it if
>> you don't use the word "SD" in any marketing, advertising, or
>> documentation for your product but that seems a dishonest way to do
>> business.
>>
>> //Jesse Off
>>
>>
>
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/ts-7000/
<*> Your email settings:
Individual Email | Traditional
<*> To change settings online go to:
http://groups.yahoo.com/group/ts-7000/join
(Yahoo! ID required)
<*> To change settings via email:
<*> To unsubscribe from this group, send an email to:
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|